Let us begin with answering what determinism is, and then turn to two notions of free will, one that is compatible with determinism and one that is not. I think you’ll find that the version of free will that is incompatible with determinism lines up better with our common or everyday notion of free will.
Determinism is the view that the state of the world at an earlier time determines or is causally sufficient for the state of the world at all later times. The state of the world is understood to pick out all features of the world and interactions in it. For example, under determinism, a westerly wind hitting the sails of a boat in just the way it does with the boat weighing what it does and the ocean giving it the buoyancy it has and so forth is sufficient for the sail boat to change its trajectory in just the way it does. Here we have a world state, the totality of the facts surrounding the boat, being sufficient for a subsequent world state, roughly all those same facts except the updated facts concerning the trajectory of the boat.
Sometimes laws of nature are included in deterministic formulations. For these, we can simply update our formulation to say that the state of the world at an earlier time, alongside the laws of nature that operate in or over the state of the world, determines or is causally sufficient for the state of the world at all later times. (As an aside, I find the notion of laws of nature to be redundant in that a full description of the state of the world will include a description of how things causally interact with one another; furthermore, I would not be sympathetic to the thought that laws of nature stand apart from the things themselves to guide them from the outside as if such laws were things in the world that causally interacted with things that are not laws. In short, we should not in general view things as being, on the one hand, some object, and, on the other, a law of nature that informs how it behaves. That is, a singular thing is not plural.)
Another way of stating determinism is with my own pet formulation: Given the position of all things and their nature and interactions with one another, nothing about their future position and nature is unsettled.
Under such formulations, indeterminism, as laid out in my Indeterministic Causal Relations post, is barred. Recall that an indeterministic causal relation has to do with the effects of a cause being probabilistic, in that one and only one of its possible effects will occur. Being probabilistic means that each of these possible effects has a probability value where the values of each and every possible effect adds up to one, and the total probability space has to do with one and only one possible effect actually occurring.
The reason that such a picture is barred under determinism is that including an indeterministic cause of this sort in the world state would indicate that the world state is insufficient for any particular future effect. In other words, the facts of the current world state would not be enough to arrive at all the future facts of the world state, specifically where the effects are probabilistic. Of an indeterministic cause, we can ask “does one effect or does another effect occur in the future, from the current causes?” Such effects will have to be brought out in the future, and are not currently available or set by the present items. Such facts belong to the future and are not contained in the present, aside from the fact that one of the cause’s effects in the set of possible effects will occur to the exclusion of the others.
Compatibilist Free Will
Under Compatibilism, free will is often put in terms of wanting or desire. You are free to do something if you want to do it, and if your desire were different, you would want, and thus do, something else. This isn’t to say that there is a guarantee of doing whatever you want, but just that the change of desire would change what you are currently doing. Perhaps you’d simply end up not doing whatever you are initially doing, given the new desire. The compatibilist notion is more generally put in the form of a counterfactual: had you wanted to do something else, you could have done so. The sense of ‘could’ here is just that wanting differently is sufficient for acting differently, for if it were not, then you were not really free to act differently. Other forms of compatibilism also fit in reasons for action. So being free to act can also indicate that had you reasoned to act differently, you could have done differently. Yet another form may involve acting based on what seems best. So whatever the basis for action is, in general the compatibilist can use that term that is the basis for action in the form of: had your [term for basis for action: e.g. desires, reasons, experience of what seems best] been different, you could have done differently. Such is the notion of free will for the compatiblist in general.
Notice that the compatibilist is compatible with determinism (indeed this is where their name comes from). This is because we are often free in the sense of doing what we want or reason to do, and this is true even if our desires and reasons are entirely determined by prior causes. All that matters under compatibilism is that what we decide to do lines up with what we want or reason to do. If desires or reasons lining up with actions were all there were to our notion of free will, then free will and determinism would indeed be fully compatible.
Libertarian Free Will
Unfortunately for compatibilism, there is more to our notion of free will than just desires or reasons lining up with action. For example, notice how we often desire incompatible courses of action when we are feeling indecisive. Do I eat pizza or cheeseburgers? Do I wait a little longer before eating anything? I can have desires (and reasons) for each of these choices and more. How might the compatibilist handle competing desires and reasons?
It would seem the only recourse for the compatibilist is to say that, no matter what, the most desirous choice, or the seemingly best reason, wins in the end. Such a position is not always plausible. For example, sometimes people act irrationally and against one’s most desirous choice, yet this should not automatically indicate that such a course of action is not free. Indeed, it may seem to be revealing of an underlying freedom.
The more prominent problem here for the compatibilist is the consideration that free will does not always depend on desires or reasons at all. Such things are thought to merely influence decision rather than strictly determine them.
The thought behind free will is not to be that decisions are without desires or reasons but that, given the same set of reasons and desires, the decision nevertheless could have turned out differently. Such is my formulation of free will, one we might consider to be of the libertarian view of free will. To put it another way, where the formulation is explicitly anti-compatibilist-free-will: we sometimes come to think after a choice, that despite choosing one thing, we could have chosen something else, everything else being as they were.
I strongly suspect that such a formulation fits in well with what one has in mind when speaking about acting of one’s own free will.
It is difficult to see how such a formulation of free will could be compatible with determinism. Since the state of the world is sufficient for all subsequent states of the world under determinism, the libertarian view of free will holds that choices that affect the world state differently could be, regardless of any change in the world state. That is, some world state W may precede some choice C which affects the world state W to be a new one, CW. Yet the same world state W may also precede a choice inconsistent with C, C2, which would affect W to be the new world state C2W. So W would not be sufficient for CW, particularly when C2W is what follows (or for C2W when CW is what follows).
The sort of causation just delineated is possible, despite being typically thought of as lacking any sort of coherent causal model. In my next post, I plan to explicitly model the form of this causation involved. indeed, you may already be able to see how we might build on our indeterministic model of causation to model libertarian free will.